Cognition the Case Against Core

Cognition

The Case against Core Knowledge

The field of developmental psychology is in a state of flux over the issue of cognition and its emergence in human beings. Studies into the acquisition of knowledge have increasingly driven a wedge between researchers who argue that environmental inputs and learning are key to cognition, while others make the case for the existence of core knowledge. The core knowledge theorists argue convincingly that basic first knowledge in several realms of cognition -- such as spatial reasoning -- is hard-wired into the human brain. The suggestion is that this knowledge is innate, not environmentally learned, and therefore transmitted genetically (Lecuyer, 2006; Spelke and Kinzler, 2007). These theorists rely predominantly on studies with infants to describe what they believe is innate, core knowledge that is universal in the human species.

However convincing or appealing this argument might be, it is based on dubious logic and assumptions. The core knowledge theorists make a series of leaps of logic that border on the irrational. Included in these leaps is the willingness to include human infants in a class of being all their own, alongside human adults and various species of primates. In addition, based on their growing body of research, these researchers are more than willing to ignore or sideline the wealth of literature that already exists on cognitive development, effectively ignoring research that points to the importance of environmental inputs in cognitive development. Too, the conclusions drawn and the assumption upon which they are based are dubious at times, and forego simpler explanations for the observed phenomena (Newcombe, 2002; Haith, 1998).

In the end, then, we must conclude that the argument for core knowledge is currently too weak to be widely supported. Over time, as the body of research grows, this may change. For now, however, there simply is not enough evidence to support the extreme argument that human beings are born with innate, genetically-transmitted information about their environment that allows them to instantly interact with it. A much simpler and more reasonable argument is that cognition develops through environmentally learned experiences, from which the developing brain produces logical assumptions about the world and its operation.

A number of studies are used as evidence that human beings demonstrate core knowledge that must preexist before birth, even at the moment of conception written into the genetic code of the individual. The core knowledge argument is based on claims in two major areas of research: studies of early competence, particularly in spatial and quantitative reasoning; and studies that suggest the existence of preexisting modules of knowledge (Newcombe, 2002). By extension, these studies and arguments suggest that environmental experience is unimportant to cognitive development or, at least, much less important than the ephemera of core knowledge. The studies used to demonstrate the existence of core knowledge run a wide gamut, a few of which we will touch upon.

Spelke and Kinzler (2007) argue that studies of infants and non-human animals indicate there is are a few parallel domains of knowledge that appear to be present instantly and universally. The supposed universality of the knowledge examined is used to argue that it must be built into our biology, since if this kind of knowledge were learned it would manifest variations that could be tested for and identified. The capacity for language in infants is one domain that has been extensively tested for. Hespos and Spelke (2004) point out that language takes advantage of representational concepts that appear to be common in all human beings as well as in some primate species. The ability to perceive subtle auditory differences that are crucial to a language's comprehensibility are available to children, but not necessarily to adults of a different language group. The implication of this is that there are some basic capacities for language that is built into the human brain before it even hears its first language.

Similarly, Hood (2001) points out that there are numerous studies of infant perception, particularly studies that test an infant's capacity to follow linear and non-linear paths of objects in anticipation of their movement. These studies show, in general, that infants spend more time looking at these movements when they defy classic Newtonian physics. In other words, when researchers manipulate the motion of an object, say by occluding it, infants will turn their attention to these seemingly "impossible" events more so than for events that conform to everyday notions of motion. This might indicate that infants are born with the capacity to understand the physics of the world in which we all live immediately and without reservation.

Along these lines, studies have been conducted into the spatial and quantitative reasoning skills of non-Western cultures, particularly those that are extremely isolated or different from our own. Dehaene, Izard, Pica, and Spelke (2006) investigated the basic geometric knowledge of the Munduruku tribe in the Amazon. Their conclusion was that these individuals -- children and adults were tested -- were able to grasp and utilize basic geometric concepts like the line, right angles, and parallelism without a base knowledge of Euclidean geometry. The argument drawn from this conclusion was that the tribe was tapping into core knowledge of spatial reasoning that was common to all human beings and existed even before the moment of birth.

The problem with this study in particular is that it ignores the possibility that an Amazonian tribe could have deduced, if only implicitly, Euclidean principles of geometry. The implication of this research is that without the benefit of Western schooling or a Greek patriarch, it is unreasonable to think that a tribe living deep in the Amazon could have developed basic concepts such as the line or right angles. In fact, the study itself presented these concepts to its subjects in terms of nonverbal exercises that would apply to the life of the tribe, such as hunting. It is more than reasonable to assume that a tribe that derives a significant portion of its caloric intake from hunting would have refined its ability to hunt, taking advantage of tactics that envision the hunter moving parallel to his prey or setting a trap along the presumed line of the prey's motion. In fact, it is exactly this kind of knowledge that anthropologists have long known to be passed on from generation to generation via culture. The fact that this knowledge also employs geometry does not necessarily suggest that geometry is a component of core knowledge that the tribe -- and the rest of us -- was born with. Instead, it simply suggests that obvious: Euclidean geometry is a mathematical representation of the mechanics of the universe, mechanics that can be deduced by any human being, even if they cannot characterize those mechanics in mathematical terms.

The core knowledge argument makes the case for the existence of knowledge modules, like geometry or spatial reasoning, that are preprogrammed into the human brain. But the evidence and logic behind this assumption appears dubious at best. As in the example noted above, it is generally simpler to explain this phenomenon by arguing that the human brain is capable of quickly coding environmental information based on limited spatial inputs (Newcombe, 2002). In other words, human beings are capable of learning very quickly and incorporating environmental data into their behaviors at a rapid pace. So rapid, in fact, researchers might mistake it for early competence. Though it is interesting to note that research performed into early competence almost always deals with infants who are at least a few months old. Is unreasonable to assume that in those few months infants, already known to be expert learners, could have incorporated basic environmental information into their behaviors that might be mistaken for core knowledge? Any conclusion of core knowledge based on early competence could simply end up being an example of environmental learning that previously was going undetected (Newcombe, 2002).

Even studies that have been constructed in order to demonstrate the existence of core knowledge often fall short of that goal and only reinforce the criticism that claims of core knowledge are overblown at best (Haith, 1998). A study by Hofsten, Feng, and Spelke (2000) was designed to test whether or not infants were capable of extrapolating the motion of objects on linear and nonlinear paths. The study itself came to four primary conclusions. One, there was no evidence that linear extrapolation is an innate skills. Two, infants can learn to accurately anticipate linear motion. Three, infants can learn to predict how occluded objects move, but not where they will appear even when the path is linear. Four, infants were able to learn to anticipate nonlinear motion, but this learning progressed much more slowly than when the infants were dealing with linear motion. What we can draw from this is that infants are especially adept at learning how to incorporate environmental information into their cognitive processes, but are not born with any innate abilities. At least, in this case, it seems apparent that spatial reasoning abilities are not innate abilities that are part of a core knowledge repertoire. Instead, spatial reasoning appears to be based on environmental inputs and old-fashioned cognitive development.

Why this should come as such a surprise to some researchers is uncertain. Core knowledge theorists claim that infants almost immediately express certain types of knowledge. But this suggestion assumes two things: one, that it is possible to measure infant cognition at the moment of birth; and two, that infants are incapable of learning before they are born. On the matter of the former point, it seems apparent that logistical and ethical concerns would make it exceedingly difficult, if not outright impossible, to test infant cognition immediately after birth. With regards to the second issue, we already have evidence that infants are capable of basic learning while still in the womb. Though developmentally unfinished, the basic sensory organs that the fetus develops permit it to learn information about its environment. Lecuyer (2006) reminds us that it is established that infants are capable of learning before birth, in particular of distinguishing the voice of their mothers from the voices of strangers. At the same time, the brain is developing structures and mechanisms that make the acquisition of knowledge easier and more effective. Thus, rather than assuming that we are all born with core knowledge domains built into our genes, it is much more reasonable to deduce that evolution has granted human beings -- and probably many other species -- with the capacity for rapid learning and adaptation to our environment. In fact, we could posit that without such an ability, survival rates for many species would be much lower because they would be unable to process and synthesize environmental information rapidly enough to ensure survival.